APPLIED PHYSICS LETTERS
VOLUME 76, NUMBER 15
10 APRIL 2000
Temperature dependence of the Raman spectra of single-wall carbon nanotubes H. D. Li, K. T. Yue,a) Z. L. Lian, Y. Zhan, L. X. Zhou, and S. L. Zhangb) Department of Physics, Peking University, Beijing 100871, People’s Republic of China
Z. J. Shi and Z. N. Gu Department of Chemistry, Peking University, Beijing 100871, People’s Republic of China
B. B. Liu, R. S. Yang, H. B. Yang, and G. T. Zou National Laboratory of Superhard Materials, Jilin University, Changchun 130023, People’s Republic of China
Y. Zhang and S. Iijima Fundamental Research Laboratories, NEC Corporation, 34 Miyukigaoka, Tsukuba, Ibaraki 305-8501, Japan
共Received 8 October 1999; accepted for publication 14 February 2000兲 Raman spectra of single-wall carbon nanotubes 共SWCNTs兲 were measured at different temperatures by varying the incident laser power. The elevated temperature of the SWCNTs and multiwall carbon nanotubes 共MWCNTs兲 is confirmed to be due to the presence of impurities, defects, and disorder. The temperature coefficient of the frequency of the C–C stretching mode E 2g 共GM兲 and that of the radial breathing mode in the SWCNT were determined to be ⬃⫺0.038 and ⬃⫺0.013 cm⫺1/K, respectively. It is found that the temperature coefficient of the GM in the SWCNT is larger than that of the MWCNT, highly oriented pyrolytic graphite, and the graphite. This is attributed to the structural characteristic of the SWCNT—a single tubular carbon sheet with smaller diameter. © 2000 American Institute of Physics. 关S0003-6951共00兲04615-5兴
Raman spectroscopy is a nondestructive technique for structural characterization of many materials. In recent years, Raman spectra of carbon nanotubes 共CNTs兲 have attracted a lot of attention, both theoretically and experimentally.1–3 Both multiwall carbon nanotubes 共MWCNTs兲 and singlewall carbon nanotubes 共SWCNTs兲 have been studied. Temperature-induced changes in the vibrational bands of carbonaceous materials including carbon fiber,4 highly oriented pyrolytic graphite 共HOPG兲,5 disordered carbon,6 and more recently, MWCNTs,7–9 were investigated using Raman spectra. In our previous report,9 we have systematically studied the temperature effect induced by laser heating of MWCNTs and other materials such as active carbon 共AC兲, HOPG, and diamond. We found that increases of the incident laser power induced corresponding increases in the temperature of MWCNTs and AC, but not for HOPG and diamond. In addition, the Raman peak position of the optical C–C stretching mode E 2g 共GM兲 of the MWCNT and AC at ⬃1585 cm⫺1 shifts to lower frequency with increasing temperature. In this letter, we report further studies on the laserinduced heating and temperature-reduced effect of the Raman shift of CNTs and its origin to also include study on SWCNTs. The SWCNT samples were synthesized by the arcdischarge method.10–12 The transmission electron microscopy image of a purified sample is shown in Fig. 1. As seen in the image, most SWCNTs in the sample have a diameter of 1.3–1.4 nm and they tend to self-organize into bundles.
All Raman spectra were measured in a backscattering geometry by a Renishaw 1000 Raman microscope with a Linkam THMS600 heating stage. The 514.5 nm line of an Ar⫹ laser was used as the excitation. The laser power ranged from 50 W to 5 mW at the sample and was focused into a spot size of 1–2 m in diameter, giving a power density of 102 – 104 W/cm2. The sample temperature of the measured spot was determined in situ by the ratio of the intensity of the Stokes and anti-Stokes Raman peaks.9 Our previous study reported the laser-induced temperature increase in the MWCNT and AC, but not in HOPG and diamond,9 suggesting that the ability to induce the temperature increase in carbonaceous materials may originate from the presence of impurities, defects, and disorder in the samples. In order to confirm this explanation, we measured the temperature of unpurified SWCNTs, purified SWCNTs,
a兲
Also with: the College of Life Sciences, Peking University. Author to whom correspondence should be addressed; electronic mail:
[email protected]
b兲
FIG. 1. Typical high-resolution electron microscopy image of the purified SWCNTs sample.
0003-6951/2000/76(15)/2053/3/$17.00 2053 © 2000 American Institute of Physics Downloaded 28 Jun 2009 to 130.207.50.192. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
2054
Appl. Phys. Lett., Vol. 76, No. 15, 10 April 2000
FIG. 2. Temperatures calculated from Raman peaks of the unpurified, purified SWCNT, and Au–HOPG as a function of incident laser power.
and Au-ion implanted HOPG 共Au–HOPG兲 by laser-induced heating. The Au ions were implanted with a dose of 1015 ions/cm2 at an accelerating voltage of 100 keV. Clearly, ion implantation introduces impurities, defects, and disorder into the HOPG. Furthermore, the density of impurities in purified SWCNTs is lower than that in unpurified ones. Figure 2 shows the temperature increase of the samples with increasing the laser power. As seen in Fig. 2, the temperature of the Au–HOPG increases with increasing the laser power, and the same laser power induces a higher temperature in the unpurified SWCNTs than that in purified SWCNTs. Recalling that no temperature increase with increasing the incident laser power for HOPG,9 the results confirm that the ability of laser-induced heating of these carbonaceous materials depends on the presence of impurities, defects, and disorder with high density. CNTs synthesized by the arc-discharge method inevitably contain many impurities 共other carbonaceous materials, e.g., C60 , disorder carbon, etc.兲, defects 共such as some polyhedral defects in the cap of squat SWCNTs兲, and disorder 共tangling with each other兲 in the final product.10–12 The impurity, defect, and disorder presence reduces the thermal conductivity of the CNTs sample significantly, resulting in the thermal conductivity usually being much lower than that of the HOPG and diamond. The high thermal conductivity of HOPG and diamond allows the samples to dissipate laserinduced heating effectively, resulting in minimal to no increase in sample temperature. In contrast, the lower thermal conductivity in CNTs makes heating dissipation inefficient. Therefore, it is expected that the temperature of CNTs samples appears to increase with increasing laser power, as shows in Fig. 2. Figure 3 shows the change of the Raman spectra of the SWCNT at different temperatures dues to laser illumination. With increasing temperature the wave numbers of the Raman peaks downshift simultaneously. Note that changes occur not only in the GM of the optical phonons, but also in the radial breathing mode 共RBM兲 of the acoustic phonons centered at about 180 cm⫺1 that is not observed in the MWCNTs. The result, together with that of the MWCNTs,9 indicates that the
Li et al.
FIG. 3. Raman spectra of the SWCNT between 150 and 225 cm⫺1 and between 1400 and 1700 cm⫺1 at temperatures of 435, 500, and 600 K, respectively.
strong temperature effect is a common property of all CNTs. It is shown in Fig. 3 that the Raman peak of optical mode GM at ⬃1585 cm⫺1 observed in the MWCNT, splits into three distinguishable peaks at ⬃1550, 1567, and 1590 cm⫺1, respectively, in the SWCNT, which was also reported in Ref. 13. In this letter, we only discuss the dominant peak at 1590 cm⫺1. The Raman peak positions of the GM and RBM of the SWCNTs as a function of temperature are shown in Fig. 4. The straight lines are linear least-square fits. Table I lists the slopes 共i.e., temperature coefficient of Raman frequency shift兲 and the intercepts of the fitted straight lines. The values of the MWCNTs,9 graphite,4 and HOPG 共Ref. 5兲 are also listed in Table I for comparison. Electric heating was also carried out in a Linkam TH600 heating stage to study the effect of temperature on the Raman spectra of the CNTs. The results are presented in Fig. 4 and listed in Table I. As shown in Fig. 4 and Table I, the downshifts of the GM and RBM frequencies induced by electric heating are basically identical to those by the laser illumination. It indicates that the frequency change of GM and RBM under laser illumination originates from the laser-induced heating effect.
FIG. 4. Temperature dependence of the frequencies of the GM and RBM for SWCNTs heated by laser 共marked by the symbol ‘‘䉱’’ and solid fitted lines兲 and by a Linkam TH600 stage 共marked by the symbol ‘‘䊊’’ and the dashed fitted line兲. Downloaded 28 Jun 2009 to 130.207.50.192. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
Li et al.
Appl. Phys. Lett., Vol. 76, No. 15, 10 April 2000
2055
TABLE I. Slopes and intercepts of the fitted straight line for the GM and RBM of the SWCNT in Fig. 2 and the reported results of the MWCNT, graphite 共Lonza Ks-5-75兲 共GL兲, and HOPG. SWCNTa Sample
Laser
Slope (10⫺2 cm⫺1/K)
GM RBM
Intercept (cm⫺1)
GM RBM
Heating stage
⫺3.8 ⫺1.3 1606 193
⫺4.2 ⫺1.5 1609 192
MWCNTb ⫺2.8
GLc
HOPGd
⫺3.0
⫺2.8
1611
a
Present work. Reference 9. c Reference 4. d Reference 5. b
According to the temperature coefficients of the GM frequencies shown in Table I, the value of the SWCNT is significantly larger than that of the MWCNT, HOPG, and graphite, and the latter have a basically identical value. As discussed in previous literature, the frequency downshift of the GM in the above carbonaceous materials is associated with the lengthening of the C–C bond.4–9 The longer bond length results from the thermal expansion of the material upon the temperature increase. To seek the origin of the different temperature coefficients, we think that it is easier to lengthen the length of the C–C bonds for the SWCNT than for the MWCNT, graphite, and HOPG under the same changing temperature. It may be attributed to the difference in structures among these carbonaceous materials. An ideal SWCNT is formed by a planar graphene sheet rolled up into a seamless, cylindrical tube. Because of the curvature, when the C–C bonds on the cylindrical tube stretch the restoring force will have a new radial component, which is proportional to the curvature of the cylindrical tube, except that it is in the direction of the C–C bond. As a result, the expansion in the tangential direction of the SWCNT is weakly restricted compared to that in planar graphite and one would expect the C–C bond of the SWCNT to be easily expanded with increasing temperature. As for the GM, which is due to the displacement of adjacent carbons in the C–C bond direction, it is anticipated that the curved SWCNT should have a larger temperature coefficient for the GM shift than the planar carbon materials, such as the graphite and HOPG. Furthermore, the expansion of C–C bonds will induce an increase of SWCNT diameter upon increase in temperature. Then, it should also be expected to have a higher temperature coefficient of the RBM, which is due to the vibration in the radial direction. Since MWCNTs are a multiple-layer structure, and have an inner and outer diameter range of 10–50 nm,14,15 one order of magnitude larger
than the diameter 共1–2 nm兲 of SWCNTs, the structural effects on the vibration properties for MWCNTs are much closer to those of planar graphite than to those of SWCNTs, although both the SWCNT and MWCNT are tubular materials. In fact, our results in Table I show that the curvature effect MWCNT is negligible, giving a temperature coefficient comparable to that of graphite. This work is supported by the State Committee of Science and Technologies, the National Science Foundation of China 共NSFC Nos. 19774009 and 69876001兲, and the China Postdoctoral Science Foundation. 1
R. A. Jishi, L. Venkataraman, M. S. Dresselhaus, and G. Dresselhaus, Chem. Phys. Lett. 209, 7 共1993兲. 2 A. M. Rao, E. Richter, S. Bandow, B. Chase, K. A. Williams, S. Fang, K. R. Subbaswamy, M. Menon, A. Thess, R. E. Smalley, G. Dresselhaus, and M. S. Dresselhaus, Science 275, 187 共1997兲. 3 H. Hiura, T. W. Ebbesen, K. Tanigaki, and H. Takahashi, Chem. Phys. Lett. 202, 509 共1993兲. 4 N. J. Everall, J. Lumsdon, and D. J. Christopher, Carbon 29, 133 共1991兲. 5 A. Erbil, M. Postman, G. Dresselhaus, and M. S. Dresselhaus, Extended Abstracts and Program of the 15th Bienn. Conference of Carbon, Pennsylvania 共1981兲, p. 48. 6 D. B. Fischbach and M. Couzi, Carbon 24, 365 共1986兲. 7 R. S. Ruoff and D. C. Lorents, Carbon 33, 925 共1995兲. 8 P. V. Huong, R. Cavagnet, P. M. Ajavan, and O. Stephan, Phys. Rev. B 51, 10048 共1995兲. 9 F. M. Huang, K. T. Yue, P. H. Tan, S. L. Zhang, Z. J. Shi, X. H. Zhou, and Z. N. Gu, J. Appl. Phys. 84, 4022 共1998兲. 10 Z. J. Shi, Y. F. Lian, X. H. Zou, Z. N. Gu, Y. G. Zhang, S. Iijima, L. X. Zhou, K. T. Yue, and S. L. Zhang, Carbon 37, 1449 共1999兲. 11 Y. Zhang, S. Iijima, Z. Shi, and Z. Gu, Philos. Mag. Lett. 79, 473 共1999兲. 12 B. Liu, T. Wagberg, E. Olsson, R. Yang, H. D. Li, H. B. Yang, S. L. Zhang, G. T. Zou, and B. Sundqvist, Chem. Phys. Lett. 共accepted兲. 13 A. Kasuya, Y. Sasaki, Y. Saito, K. Tohji, and Y. Nishina, Phys. Rev. Lett. 78, 4434 共1997兲. 14 Z. J. Shi, X. Zhou, Z. Jin, Z. Gu, J. Wang, S. Feng, X. Xu, and Z. Lin, Solid State Commun. 97, 371 共1996兲. 15 C. H. Kiang, M. Endo, P. M. Ajayan, G. Dresselhaus, and M. S. Dresselhaus, Phys. Rev. Lett. 81, 1869 共1998兲.
Downloaded 28 Jun 2009 to 130.207.50.192. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
